NB
Back to Selected Projects
AIStatus: ActiveYear: 2026

AI Study Coach

An intelligent study assistant transforming raw lecture notes into active recall flashcards and quiz evaluations.

01. Overview & Description

Created to streamline my own CS exam preparation across operating systems, database management, and computer networks.

02. System Architecture Diagram

Web UI -> FastAPI Parser -> Local LLM JSON Validator -> Spaced Repetition Engine

Interactive System Flow Diagram

Click any node below to inspect its operational role in the pipeline.

● Live Animated Flow
1React18 DashboardFrontend Client2FastAPIEngineREST & SSE Middl3Authentication/ RLSSecurity Policy4PostgreSQL/ SQLitePersistence Laye5LocalOllama LLMQuantized Infere6StreamedResponseUI Rendering Out
Selected Node: React 18 Dashboard(Frontend Client)

Handles user prompt input, interval charts, SSE streaming data rendering, and dark-theme state management.

03. Database Schema & API Specifications

Database Schema Design

Local storage index storing flashcard decks, SRS review intervals, and historical quiz scores.

API Endpoints
  • POST /api/generate-quiz
  • POST /api/evaluate-answer

04. Codebase Folder Structure

ai-study-coach/
├── backend/
│   ├── app/parser.py
│   └── app/prompt_templates.py
├── frontend/
│   └── src/components/QuizCard.tsx
└── README.md

05. Visual Evidence & Artifacts

Interactive Quiz Interfacecode

$ ai-study-coach --status active

[OK] System pipeline initialized

Instant active-recall feedback on CS concept notes

06. Engineering Challenges

Enforcing deterministic JSON schema outputs from open-source LLMs without hallucinated keys.

07. Key Lessons Learned

Prompt validation and JSON schema enforcement are essential when relying on local open-source language models.

Technologies Used

AIPythonFastAPIReactTypeScriptOllama